In the present project, statistical and epidemiologic methods are developed to facilitate identification of genetic determinants of complex traits. Theoretical and simulation-based approaches are used to evaluate the properties of these methods in various situations. The methods are applied to data generated in other projects, particularly "Genetic Epidemiology of Diabetes and Obesity" (Z01 DK069028) to help assess the genetics of type 2 diabetes and related traits.[unreadable] In the past year methods were developed for summarizing the power of a map of markers for genetic association studies and these were applied to a fine-mapping project of obesity, which has shown linkage to chromosome 11q. In addition a method for assessing whether an association can account for a linkage result was developed. In application to an ongoing mapping project for young-onset type 2 diabetes on chromosome 1q, this method suggested that association of markers in the CACNA1E gene may explain a significant proportion of the linkage signal. A method that combines general association statistics (which are powerful but potentially confounded by population stratification) with within-family association tests (which are less powerful but robust to stratification) was developed in an attempt to retain the desirable characteristics of both tests. This test was applied to a genome-wide association study of young-onset type diabetes involving 100,000 makers. Several regions with potential diabetes-susceptibility genes were identified. Marker panels for mapping by admixture linkage disequilibrium in populations of mixed European and Amerindian origin have also been developed.[unreadable] Current efforts involve additional application of the tests that combine evidence from general association, within-family association and linkage tests to data from genome-wide association studies and other mapping projects. It is hoped that these methods will facilitate the identification of potentially important variants for further functional study. In addition path analytic methods are being evaluated to examine pleiotropic associations of genetic polymorphisms with diabetes and it quantitative subphenotypes. These are being applied to studies of diabetes and measure of insulin secretion, insulin sensitivity and obesity. Methods are also being developed for analysis of genetic transcription data that account for alternate splicing patterns. These are being applied to data from an exon-specific array applied to skeletal muscle tissue to identify transcription patterns that predict development of type 2 diabetes.